Possible to Incrementally track data in database?

[jon49]

The database I have is small and I need to track the incremental changes. I would prefer not to do full back ups every time. Is it possible to do incremental git changes programmatically?

Thanks!

[anupsaund]

Hi Jon,

This is a fantastic idea!

We ran some tests to see how we would go about implementing this feature and we have come up with the following workflow that may work.

Export

1. Record a date and time of the first full export e.g. { AdventureWorks2012 : 14/10/2016 09:00:00}
2. Offer a button on the 2nd page, to Show changes since last update. (This button would be greyed out or hidden until a full export had been performed)
3. Pressing the Show changes since last update would then filter the list of discovered objects down to the ones that have been modified since the last full export.

Import

• We have not thought of any way of implementing an incremental import yet. Our discussions lead us to believe it may not be possible.

Have you got any thoughts about an incremental import? Would it be a useful feature?

Let me know what you think of the Export flow - and if it meets your requirements.

[flippidippi]

This is something we would want also. And icing on the cake would be if we could automated this process at the end of each day to catch all changes. This would help us move away from using redgate DLM to gitsql.

[jon49]

Would there be a way to programmatically do the export? So, each day I could just run a script that does the export. Or even after I do an insert/update I send a command to sync with git. Then I could do a commit message too with whatever change was made.

I don't think I would need to do an incremental import. It might be nice but not necessary. The only time I would get the data back would be if we need to recreate the database from a certain time frame. Let's say someone is suing us and they said that they got the information from us on a particular date. Then we could recreate that day and see exactly what was given as advice to the client.

The writes aren't really that often. Maybe a 10 times a day during inactive periods and during active periods quite a bit more. And I would only do the backups when an update is made that would affect release.

Anyways, that's more of our use case.

[anupsaund]

Hi @jon49 and @flipxfx

We could add an option in as part of the #3 Command Line Options feature.
The option could then be used to automate the process via a scheduled job on windows scheduler.

As an example, one schedule in the morning, and one in the evening, with the ability to run it ad hoc during the day.

Thank you for the feedback, we would love to get this into the next release for you.
At the very least, the incremental export feature, on the GUI, then eventually as the CLI option.

[sandeep0308]

Hi All
I am looking for detail documentation on Gitsql source control to maintain the history of the database objects as and when updated/deleted. Do Gitsql have all the features of source control like TFS or red gate? I would like to know how the version numbers can be maintained using it? I would appreciate if someone throw some light on this? Thanks! Have a great day!

[mariascott]

@sandeep0308 documentation can be found at https://gitsql.net/documentation#SQLServer

[sandeep0308]

Thank you! I went through this documentation and yet to get clarification on how to maintain the versions and history in a repository, which was missing in the documentation part. All I could notice is import and export of DB objects, but there is some link missing about how to view the previous changes and files that are updated earlier. Please suggest. From: Maria Scott [mailto:notifications@github.com] Sent: Tuesday, September 19, 2017 3:22 PM To: gitsql/SQL-Server-Edition <SQL-Server-Edition@noreply.github.com> Cc: Sandeep Nagulapally <Sandeep.Nagulapally@surya.com>; Mention <mention@noreply.github.com> Subject: Re: [gitsql/SQL-Server-Edition] Possible to Incrementally track data in database? (#20) @sandeep0308<https://github.com/sandeep0308> documentation can be found at https://gitsql.net/documentation#SQLServer — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub<#20 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AenNxOUKcuP7SteUUyirV8m9g5MCee-Nks5skBRdgaJpZM4KU7X1>.

[mariascott]

@sandeep0308 - I see what you mean...

It really depends on the source control software which is being used to how files are to be version controlled. gitSQL aims to deliver flat files of data and schema to the filesystem - to then be used by version control software of choice to compare/merge conflicts/commit into source control.

What do you use for version control?

Perhaps we can look at putting together some tutorials for specific use cases?

[sandeep0308]

I want to use Gitsql over Redgate, which is cost effective. I see a comparision chart in your website. But gitSql doesn’t maintain the history when I try doing import/export of Database objects. https://gitsql.net/gitsql_vs_redgate From: Maria Scott [mailto:notifications@github.com] Sent: Tuesday, September 19, 2017 3:42 PM To: gitsql/SQL-Server-Edition <SQL-Server-Edition@noreply.github.com> Cc: Sandeep Nagulapally <Sandeep.Nagulapally@surya.com>; Mention <mention@noreply.github.com> Subject: Re: [gitsql/SQL-Server-Edition] Possible to Incrementally track data in database? (#20) @sandeep0308<https://github.com/sandeep0308> - I see what you mean... It really depends on the source control software which is being used to how files are to be version controlled. gitSQL aims to deliver flat files of data and schema to the filesystem - to then be used by version control software of choice to compare/merge conflicts/commit into source control. What do you use for version control? Perhaps we can look at putting together some tutorials for specific use cases? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub<#20 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AenNxMkdRydhNhw6-3Cv319JPBMMXKjtks5skBkMgaJpZM4KU7X1>.